Cholinergic medication for neuroleptic-induced tardive dyskinesia

Drug-induced tardive dyskinesia is a common adverse effect of some antipsychotics, especially when these are given for an extended period of time. Tardive dyskinesia consists of involuntary repetitive movements, mainly in the oral region, but sometimes also in the limbs. It may become persistent. Cholinergic drugs, such as deanol, lecithin and meclofenoxate, have been used to treat tardive dyskinesia. This review did not identify any evidence to suggest that they are effective and found some to suggest that these old drugs may be toxic. New cholinergic drugs have been developed for the treatment of Alzheimer's disease, and it will be of interest to know if these drugs have an effect on the movements of tardive dyskinesia. We found one ongoing randomised trial.

Authors' conclusions: 

Tardive dyskinesia remains a major public health problem. The clinical effects of older cholinergic drugs are unclear, as too few, too small studies leave many questions unanswered. Cholinergic drugs should remain of interest to researchers and currently have little place in routine clinical work. However, with the advent of new cholinergic agents now used for treating Alzheimer's disease, scope exists for more informative trials. If these new cholinergic agents are to be investigated for treating people with tardive dyskinesia, their effects should be demonstrated in well-designed, conducted and reported randomised trials.

Read the full abstract...

Tardive dyskinesia remains a troublesome adverse effect of conventional antipsychotic (neuroleptic) medication. It has been proposed that tardive dyskinesia could have a component of central cholinergic deficiency. Cholinergic drugs have been used to treat tardive dyskinesia.


To determine the effects of cholinergic drugs (arecoline, choline, deanol, lecithin, meclofenoxate, physostigmine, RS 86, tacrine, metoxytacrine, galantamine, ipidacrine, donepezil, rivastigmine, eptastigmine, metrifonate, xanomeline, cevimeline) for treating neuroleptic-induced tardive dyskinesia in people with schizophrenia or other chronic mental illness.

Search strategy: 

An electronic search of the Cochrane Schizophrenia Group's register (October 2001) was undertaken. This register is assembled by extensive searches for randomised controlled trials in many electronic databases, registers of conference proceedings and dissertations. References of all identified studies were searched for further trial citations. Principal authors of trials were contacted.

Selection criteria: 

Reports identified by the search were included if they were of controlled trials dealing with people with neuroleptic-induced tardive dyskinesia and chronic mental illness, who had been randomly allocated to either a cholinergic agent or to a placebo or no intervention. Two reviewers independently assessed methodological quality of trials.

Data collection and analysis: 

Two researchers extracted data and, where possible, estimated relative risks (RR) or weighted mean differences (WMD), with 95% confidence intervals (CI). Data were analysed on an intention-to-treat basis, with the assumption that people who dropped out had no improvement.

Main results: 

We included eleven studies investigating the use of older cholinergic drugs compared with placebo. Most studies involved small numbers of participants (5-20 people). We found no completed trials of the new cholinergic Alzheimer drugs for the treatment of tardive dyskinesia.
Cholinergic drugs did not result in any substantial improvement in tardive dyskinesia symptoms when compared with placebo (8 RCTs, 170 people, RR no important improvement 0.84 CI 0.68 to 1.04). Neither did tardive dyskinesia symptoms increase (7 RCTs, 137 people, RR deterioration in tardive dyskinesia 1.17 CI 0.55 to 2.50). Pooled results for endpoint AIMS scores were equivocal (4 RCTs, 86 people, WMD -0.19 CI -0.53 to 0.14). Deanol may cause gastric adverse effects (5 RCTs, 61 people, RR 9.00 CI 0.55-148) and other adverse effects such as sedation and peripheral cholinergic effects (6 RCTs, 94 people, RR 6.83 CI 0.99-47). One study reported on global outcome. Meclofenoxate was neither clearly helpful nor harmful when compared with placebo (1 RCT, 60 people, RR not of global benefit 0.89 CI 0.59 to 1.32). We found no difference between people allocated cholinergics and those given placebo for the outcome of leaving the study before completion (10 RCTs, 240 people, RR 0.52 CI 0.21 to 1.33).